Local, wild-collected seeds of native plants are recommended for use in ecological restoration to maintain patterns of adaptive variation. However, some environments are so drastically altered by exotic, invasive weeds that original environmental conditions may no longer exist. Under these circumstances, cultivated varieties selected for improved germination and vigor may have a competitive advantage at highly disturbed sites. This study investigated differences in early establishment and seedling performance between wild and cultivated seed sources of the native grass, Poa secunda, both with and without competition from the invasive exotic grass, Bromus tectorum. We measured seedling survival and above-ground biomass at two experimental sites in western Montana, and found that the source of seeds selected for restoration can influence establishment at the restoration site. Cultivars had an overall advantage when compared with local genotypes, supporting evidence of greater vigor among cultivated varieties of native species. This advantage, however, declined rapidly in the presence of B. tectorum and most accessions were not significantly different for growth and survival in competition plots. Only one cultivar had a consistent advantage despite a strong decline in its performance when competing with invasive plants. As a result, cultivated varieties did not meet expectations for greater establishment and persistence relative to local genotypes in the presence of invasive, exotic species. We recommend the use of representative local or regional wild seed sources in restoration to minimize commercial selection, and a mix of individual accessions (wild, or cultivated when necessary) in highly invaded settings to capture vigorous genotypes and increase the odds native plants will establish at restoration sites.